Alzheimer's Reshapes How DNA Folds Inside Brain Cells, Single-Cell Study Finds

Alzheimer’s disease does not just change which genes are active in brain cells — it appears to change how the DNA itself is folded, according to a single-cell study published in Science.

Why folding controls gene activity

Two metres of DNA fit inside a nucleus a few microns across, and the arrangement is not random compaction.

The genome is organised into loops and domains that bring distant sequences into physical contact. That matters because regulatory elements controlling a gene are frequently located far away along the linear sequence — sometimes hundreds of thousands of bases distant — and can only act by being brought physically adjacent through folding.

Boundaries between domains constrain which elements can reach which genes. A regulatory element and a gene in separate domains generally cannot interact, however close they are in sequence terms.

Which means the three-dimensional arrangement is a regulatory layer in its own right: the same DNA sequence produces different gene activity depending on how it is folded.

What changed

Researchers found that in Alzheimer’s the folding is reorganised: short-range interactions within the genome decrease while longer-range ones increase, disrupting normal gene regulation.

That direction of change is informative. Short-range contacts largely represent the local domain structure that keeps regulation contained and specific. Long-range contacts connect more distant regions.

A shift from local to distant suggests loosening of the boundaries that normally partition the genome — which would allow regulatory elements to act on genes they should not reach, producing inappropriate activity rather than simple loss of function.

How they measured it

The team used GAGE-seq to measure genome folding and gene activity in the same individual cells, combined with spatial mapping and an AI model named Hicformer, applied to postmortem prefrontal cortex tissue from people with and without Alzheimer’s. The work came from Carnegie Mellon University, the University of Pittsburgh and the University of Washington.

Measuring both in the same cell is the methodological advance. Genome folding has traditionally been measured across populations of cells, producing an average that obscures variation — and brain tissue contains many cell types whose folding differs.

Averaging across a mixture where disease affects some types more than others could produce apparent changes that reflect shifting cell proportions rather than altered folding within any cell.

“The genome’s 3D structure is a fundamental regulatory layer that helps to connect DNA sequence to gene activity,” said Carnegie Mellon’s Jian Ma. Colleague Yang Zhang added that measuring both in the same cell “allows us to directly connect chromosome structure” to disease.

Why the prefrontal cortex

The region is affected later in Alzheimer’s than the medial temporal structures where pathology begins, and it is where much of the cognitive impairment that defines the disease clinically arises.

Executive function, working memory, planning and judgement depend on it, and their loss is what makes the condition disabling beyond memory failure alone.

The interpretation problem

The findings point to genome folding as a previously underappreciated regulatory layer in Alzheimer’s.

Whether the reorganisation drives disease or results from it is not resolved by this design, and the question is genuinely open. Cellular stress, altered metabolism and changes in the proteins that organise chromatin could all reshape folding as a consequence of a cell being unwell.

Postmortem tissue captures the end state, decades after whatever initiated the process, which makes distinguishing cause from consequence particularly difficult here.

What would follow

This is early, mechanistic science — not a treatment — but understanding how the disease disrupts genome architecture could reveal new targets.

Where the field has been looking instead

Genome architecture is a late addition to a list of regulatory layers Alzheimer’s research has worked through, and the sequence explains why it is being described as underappreciated.

Genetic association studies came first, identifying variants raising or lowering risk. Most of them fall outside protein-coding regions, in stretches of DNA presumed regulatory — which was informative and left the mechanism unexplained.

Gene-expression studies followed, cataloguing which genes are more or less active in diseased brains. Single-cell methods refined that to individual cell types, revealing that microglia, astrocytes and neurons change in different directions.

Epigenetic studies added chemical marks on DNA and histones, another layer governing accessibility.

Each answered part of the question and none closed the gap between a risk variant in a non-coding region and a gene whose activity changes. Three-dimensional folding is precisely the missing link in that chain: it determines which non-coding regions can physically reach which genes — which is why mapping it addresses a specific unanswered question rather than adding another descriptive layer.

Proteins that establish and maintain folding are known and in principle druggable, so an intervention is conceivable if the change proves causal. Research news, not medical advice.